Environ. Sci. Technol. 2009, 43, 7803–7809
Dynamic Cytotoxic Response to Microcystins Using Microelectronic Sensor Arrays D O R O T H Y Y U H U A N G , * ,† MELISSA MOCK,‡ BRUNO HAGENBUCH,§ SIU CHAN,| JASNA DMITROVIC,⊥ STEPHAN GABOS,# AND DAVID KINNIBURGH† Alberta Centre for Toxicology, Department of Physiology & Pharmacology, Faculty of Medicine, University of Calgary, Calgary, Alberta, Canada T2N 4N1, Department of Physiology & Pharmacology, University of Calgary, Calgary, Alberta, Canada T2N 4N1, Division of Clinical Pharmacology and Toxicology, University Hospital, Zurich, Switzerland, Department of Pharmacology, Toxicology and Therapeutics, University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas 66160-7417, Canadian Food Inspection Agency, 3650 36th Street N.W., Calgary, Alberta, Canada T2L 2L1, and Health Surveillance, Alberta Health and Wellness, Edmonton, Alberta, Canada T5J 2N3
Received April 30, 2009. Revised manuscript received August 13, 2009. Accepted August 19, 2009.
Microcystins are bioactive metabolites produced by cyanobacteria in water. These cyclic heptapeptides have caused public health concern worldwide. By interfering with cellular phosphorylation and signaling, microcystins can cause acute and chronic liver diseases. Therefore, the World Health Organization (WHO) has set the provisional drinking water guideline value at 1.0 µg/L for microcystin-LR (free plus cellbound). Microcystins do not readily cross cell membranes in in vitro cell-based assays, except for those using freshly isolated hepatocytes. However, the sensitivity of in vitro cellbased assays is not adequate for testing samples at low environmental concentrations. Hence, there is a need to develop a sensitive and stable cytotoxicity assay for use in environmental studies. On the basis of the observation that microcystin-LR can be transported by the liver-specific members of the organic anion transporting polypeptides (OATPs), we investigated the potential of using an OATP1B3-expressing cell line in a cytotoxicity assay for microcystins. Using a novel cell electronic sensing system (RT-CES), we were able to monitor the real-time, dynamic cytotoxic response to microcystins at microgram per liter concentrations. We demonstrated that the cytotoxicity of the most common microcystins, -LR, -YR, -RR, -LF, and -LW, was mediated by OATP1B3 transporters. Microcystin-LF is the most potent toxin among the five congeners. In conclusion, we have established a highly automated, real-time, sensitive, and stable assay for measuring microcystin cytoxicity. * Corresponding author phone: (403) 2205511; fax: (403) 2702964; e-mail:
[email protected]. † Alberta Centre for Toxicology, Department of Physiology & Pharmacology, University of Calgary. ‡ University Hospital. § University of Kansas Medical Center. | Department of Physiology & Pharmacology, University of Calgary. ⊥ Canadian Food Inspection Agency. # Alberta Health and Wellness. 10.1021/es901189c CCC: $40.75
Published on Web 09/17/2009
2009 American Chemical Society
Introduction Toxic cyanobacterial blooms have caused water quality and public health problems for centuries (1). The first scientific report was described by Francis in Nature in 1878 (2). The most disastrous incidence of cyanobacteria poisonings occurred in a hemodialysis unit in Brazil in 1996, where at least 47 deaths were reported (3). A group of cyclic heptapeptides known as microcystins are the major cyanobacterial toxins that have caused illness and death. Over 80 structural variants of microcystins have been characterized (4). All microcystins have a common cyclic heptapeptide structure of cyclo-D-Ala1-X2-D-MeAsp3-Z4Adda5-Glu6-Mdha7, where X and Z are variable amino acids that determine the name of the variant (5). Microcystins concentrate in the liver as a result of active uptake (6). Acute poisoning causes severe liver damage (7). Chronic low-level exposure is related to carcinogenesis and tumor promotion (8). Therefore, the World Health Organization (WHO) has set the provisional drinking water guideline value at 1.0 µg/L for microcystin-LR (free plus cell-bound), while new data are being generated to revise the guideline to include additional microcystin variants (9). Toxicity assays have been developed for microcystins. The mouse in vivo assay is a standard method employed for determining toxicity but involves animal maintenance, which is not available at most water testing facilities. Also, this assay can only be used for testing cyanobacterial blooms and not for finished water because it does not have the sensitivity required by the WHO Drinking Water guideline. Attempts have also been made to develop in vitro assays utilizing a number of immortalized cell types. The assays sensitivity (10 mg/L) is disappointing because cultured cell lines do not actively take up microcystins (10). Therefore, there is a demand for a simple, sensitive, and stable in vitro system to measure microcystin cytotoxicity. Organic anion transporting polypeptides (OATPs) are a group of carriers with a wide spectrum of amphipathic substrates (11). OATPs are expressed in multiple tissues and as a result are integral for contributing to the general absorption, distribution, and excretion of drugs and toxins. OATP1A2, OATP1B1, and OATP1B3 were demonstrated to transport microcystin-LR in a Xenopus laevis oocyte expression system (12). OATP1B3, which is exclusively expressed in the liver and certain gastric, colon, and pancreatic cancer cell lines (13), was the most active among them. Therefore, it was chosen as a candidate for establishing an in vitro testing model for microcystins. To develop a sensitive and reliable real time assay, we also incorporated a novel “cell chip” technology. This technology, known as real-time cell electronic sensing system (RT-CES, ACEA Biosciences, San Diego, CA), is based on the measurement of cell-electrode impedance (Figure 1). Microelectrodes are coated onto the bottom surface of a microtiter plate. Adherent cells are seeded into the culture wells and attach to the electrode sensor. The dynamic curve recording the impedance change indicates the physiological activity of the cells, which affects the electronic and ionic passage on the electrode sensor surfaces (14). Analog electronic readouts are measured in real time and converted to digital signals for analysis. In this study, we hypothesized that the transport of microcystins by OATP1B3 across the cytoplasmic membrane would cause cytotoxicity in cultured cells. Using a Chinese hamster ovary cell line (CHO) expressing OATP1B3, we VOL. 43, NO. 20, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7803
sensor analyzer, workstation, and 16× microtiter devices. The microelectrode sensor arrays are coated on 16× glass slides that incorporate 96 well microtiter plates (16× microtiter device). The 16× microtiter devices are connected to the workstation inside the cell culture incubator. A cable connects the workstation to the sensor analyzer for impedance measurement. The impedance data for the selected wells is exported to a computer and analyzed by RT-CES software. Cell attachment to the electrode sensor surface and morphology change will alter the local ionic environment at the electrode-solution interface, leading to a change in impedance. To quantify cell status on the basis of cellelectrode impedance, the cell index (CI) is calculated. CI )
max I)1,...,N
FIGURE 1. RT-CES technology. Cells are seeded into the plastic wells of the sensor device with 9 mm well-to-well spacing. Electronic sensors cover about 80% of the bottom surface area in each well. Devices are attached to a workstation and further connected to the sensor analyzer for impedance measurement. Measurements are sent to a computer and analyzed under the control of RT-CES software. Proliferation, apoptosis, and morphology change in the cells will increase, decrease, or transiently change the cell index, respectively. established a novel in vitro assay using the RT-CES system to measure microcystin cytotoxicity.
Materials and Methods Reagents and Media. Microcystin-LR, -YR, -RR, -LF, and -LW cell culture reagents and 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) were purchased from Sigma Chemical Co. (St. Louis, MO). Geneticin (G418 sulfate) was purchased from GIBCO (Burlington, ON). Cell Culture. Cell lines were purchased from American Type Culture Collection (ATCC) (Rockville, MD), except as noted otherwise. Mouse hepatocytes FL83B (ATCC: CRL2390) and Chinese hamster ovary cells CHO (ATCC: CCL-61) were maintained in Ham’s F12K medium with 2 mM of L-glutamine adjusted to contain 1.5 g/L of sodium bicarbonate. G418 sulfate was added to the CHO/OATP1B3 culture to a final concentration of 500 µg/mL. Stable Transfection of CHO Cells with OATP1B3. For cloning of the expression vector, we followed the procedure previously used for the expression of OATP1C1 (15). The OATP1B3 open reading frame (ORF) was PCR amplified with a forward primer containing a NheI restriction enzyme site and a reverse primer containing a NotI site. After digestion with the two restriction enzymes NheI and NotI, the resulting fragment was gel purified and ligated into a NheI/NotI cut pIRESneo2 vector (Clontech, Mississauga, ON). This resulted in directional cloning of the ORF, which was verified by sequencing. The construct was transfected into CHO cells by electroporation. Twenty-four hours after transfection the culture medium was replaced with medium containing 1 mg/mL of G418. After 10 days of culture in the presence of G418, single clones were isolated using cloning cylinders. After testing positive for transport and immunofluorescence, the cell line was recloned by limited dilution, and one of the clones, CHO/OATP1B3, was used for all future experiments. RT-CES Measurement of Cell-Electrode Impedance. There are several components in the ACEA RT-CES system (ACEA Biosciences, San Diego, CA): a computer, electronic 7804
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 20, 2009
[
]
Rcell(fi) -1 Rb(fi)
Rcell(f) and Rb(f) are the frequency-dependent electrode resistances with or without cells, respectively. N is the number of the frequency points at which the impedance is measured. The greater number of viable cells attached correlates to a higher CI value. Cell Viability Assay (MTT). Assessment of cell viability was carried out using a modified method of Mosmann (16), which is based on the transformation and colorimetric quantification of MTT as an estimate for the number of mitochondria and hence the number of living cells. Ten microliters of MTT (5 mg/mL in PBS) was added into each well for 4 h at 37 °C in a 5% CO2/95% air incubator. The medium was then replaced by 100 µL of stop solution (10% SDS in 1:1, v/v isobutanol:0.02 N HCl) overnight. The color was measured using a microtiter plate reader at the wavelength of 550 nm. Cell viability was estimated as the percentage absorbance of sample relative to the control (cells cultured without toxins). Light Microscopic Evaluation. Cell morphological changes following microcystin exposure were observed by light microscopy (Nikon, Eclipse 55i). The cell monolayer was rinsed twice with PBS and fixed in 100 µL of a fixative solution (80% methanol, 20% PBS) for 10 min at room temperature. After removing the fixative solution, cells were left to dry for 20 min before adding Giemsa stain (1:19 dilution with water). Cells were stained for 30 min then washed three times with distilled water before microscopic evaluation. Growth Curve. Cells were seeded into RT-CES 16× device wells and into corresponding 96 well culture plates at densities of 0, 1250, 2500, 5000, 7500, 10000, 15000, and 20000 cells/well. Dynamic growth curves were recorded real time by the RT-CES system. The MTT assay, an end point assay, was performed at 6, 12, 24, 30, 36, 48, and 72 h postseeding in the 96 well culture plates. Microcystin-Mediated Cytotoxicity. The assay test parameters, including cell numbers for seeding, incubation time before sample application, and culture medium components, were optimized to achieve the assay sensitivity at the microgram/liter level for microcystin -LR (data not shown). CHO/WT or CHO/OATP1B3 cells were seeded into the RT-CES 16× device wells and into corresponding 96 well culture plates at the density of 10000 cells/well. The cells were allowed to adhere for 24 h before microcystin exposure. Microcystin congeners were added to the wells at various concentrations ranging from 0.03 to 100 µg/L to induce cytotoxicity. Cell-electrode impedance was measured at 1 h intervals over 72 h using the RT-CES system, and corresponding time-dependent CI values were derived. MTT viability assay and light microscopy were conducted in parallel at 24 h after microcystin exposure to compare to the RT-CES data.
FIGURE 2. Cell number titration on the RT-CES system and with the MTT assay. (a) Growth curves of cells starting from 1250 cells/well to 20000 cells/well using the RT-CES system. Cell number titration on the RT-CES system and with the MTT assay. (b) Growth curves of cells starting from 1250 cells/well to 20000 cells/well adapted from the MTT assay at time points of 6, 12, 24, 30, 36, 48, and 72 h. Statistical Analysis. The data were analyzed with GraphPad Prism 3.0 (GraphPad Software, CA). One-way analysis of variance was used to measure the differences between selected groups.
Results Growth Curves: Data Correlation between the RT-CES System and MTT Assay. To validate the RT-CES technology, we compared the growth curves derived from the RT-CES cell index (CI, Figure 2a) and the MTT optical density (OD) readings (Figure 2b). The CI-based curves were obtained automatically by the RT-CES software, while the OD-based curves were plotted from the MTT assays performed at 6, 12, 24, 30, 36, 48, and 72 h postseeding. Both methods demonstrated cell number-dependent growth curves. By 72 h, the wells seeded with cell numbers greater than 7500 had reached the plateau phase on the RT-CES growth curves but not on the MTT curves. Taking the results at a given time point within the exponential part on the growth curve (24 h as an example), we compared the CI from the RT-CES technology and the OD from the MTT assay, both plotted against the cell numbers. Using GraphPad Prism software, a linear regression was performed on both data sets. We observed a similar goodness of fit based on the CI (r2 ) 0.98, Figure S1a of the Supporting Information) in comparison with that based on the OD (r2 ) 0.93, Figure S1b of the Supporting Information). Therefore, as an alternative to the MTT assay, the RT-CES system was used for subsequent studies to provide a realtime format for measurement of microcystin toxicity. Microcystin-LR: Not Readily Taken Up by Cultured Cell Lines. FL83B cells, a hepatocyte cell line derived from a normal mouse liver, were seeded onto the surface of RT-CES
FIGURE 3. Cytotoxic effect of microcystin-LR on cultured FL83B hepatocytes. FL83B hepatocytes were seeded at the density of 10000 cells/well and incubated for 24 h before toxin application. Different concentrations of microcystin-LR were added, and cells were incubated for up to 4 days. (a) Cytotoxicity determined by RT-CES technology. Cell index was measured every hour. Results are representative of three independent experiments. Cytotoxic effect of microcystin-LR on cultured FL83B hepatocytes. (b) Cytotoxicity determined by MTT viability assay at 48 h after exposure to microcystin-LR. * P < 0.05. 16× devices. Microcystin-LR was added at final concentrations of 0.1, 1, 10, 100, 1000, or 10,000 µg/L to induce cytotoxic responses. As shown in Figure 3a, treatment with microcystin-LR at concentrations of 100 µg/L and greater led to a reproducible decline in the cell-electrode impedance response, which indicated cell detachment and apoptosis. The microcystin-LR-mediated cell death was further confirmed by MTT assay at 48 h after toxin exposure (Figure 3b). Microcystin-LR: Causes Cell Death in CHO/OATP1B3 Cells. As shown in Figure 4a, treatment with microcystin-LR did not result in a decline in the CI of CHO/WT cells. When these cells were modified to express OATP1B3, microcystinLR exposure led to a reproducible decline in the cellelectrode impedance (Figure 4b), even at submicrogram per liter concentrations. Microcystin-LR induced cytotoxicity in CHO/OATP1B3 cells in a dose-dependent manner. Microcystin-LR at the highest experimental concentration (1000 µg/L) caused an immediate decrease in the impedance value, which reached a minimum value in 2 h. This indicated the quick rearrangement of the cytoskeleton and detachment of the cells. At lower concentrations, microcystin-LR was mildly cytotoxic, and the RT-CES system did not detect the decrease of CI until 8 h after toxin exposure. At concentrations of 0.1 µg/L and less, we observed no toxic effect throughout the experiment. The IC50 was calculated to be 1.15 µg/L at 24 h after microcystin-LR exposure. MTT assay and microscopic analysis of the cells were conducted in parallel and further supported the results from the RT-CES assay system (Figure 4c,d). Microcystin: Taken Up over a Few Hours. As shown in Figure S2 of the Supporting Information, a microcystincontaining cell culture medium was replaced by a fresh medium after certain incubation periods. The removal of microcystin-LR from the culture medium within the first 3.5 h VOL. 43, NO. 20, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7805
resulted in the absence of cytoxicity. From 4 to 9 h, partial cytotoxicity was observed. The cells could not be recovered if the medium was refreshed after 23 h of toxin exposure. Microcystin-LR, -YR, -RR, -LF, and -LW: Not the Same Toxicity on CHO/OATP1B3 Cells. There have been more than 80 microcystin variants identified (12). Each of them may have different cytotoxic potentials. We tested the most common congeners, microcystin-LR, -YR, -RR, -LF, and -LW, at various concentrations using RT-CES technology. No cytotoxicity was observed in CHO/WT cells incubated with all microcystin congeners at all concentrations tested (Figure S3a of the Supporting Information). In CHO/OATP1B3 cells, we found dose-dependent cell detachment and apoptosis caused by microcystin congeners (Figure S3b of the Supporting Information). Using 4 parameter curve fitting, 24 h IC50s were calculated as 1.03 µg/L for microcystin-LR, 2.24 µg/L for -YR, 10.07 µg/L for -RR, 0.14 µg/L for -LF, and 0.21 µg/L for -LW (Figure 5). The observations were further confirmed by MTT viability assay at 24 h post microcystins exposure (Figure S3c of the Supporting Information). Specificity of Real-Time Cytotoxicity Assay for Microcystins. To assess the specificity of this novel method, we tested selected environmental contaminants using CHO/ WT and CHO/OATP1B3 cells. To a general cytotoxic reagent such as staurosporine, we found both cell lines showed similar dose-dependent cytotoxic responses. CHO/ OATP1B3 cells selectively showed cytotoxicity to microcystins, but CHO/WT cells did not. Other common environmental contaminants, including pesticides, fire retardants, plasticizers, and endocrine disruptors, were tested, and none of the compounds showed cytotoxic results even at elevated levels (Table 1).
Discussion
FIGURE 4. Cytotoxic effect of microcystin-LR on CHO/WT and CHO/ OATP1B3 cells. Cell index was measured every hour by the RT-CES system. Results are representative of three independent experiments. (a) CHO/WT response to microcystin-LR. Cytotoxic effect of microcystin-LR on CHO/WT and CHO/OATP1B3 cells. (b) CHO/OATP1B3 response to microcystin dose-dependent cytotoxicity caused by microcystin-LR in CHO/OATP1B3 cells. (c) Cell viability based on MTT assay after 24 h microcystin-LR exposure. * P < 0.05. Dose-dependent cytotoxicity caused by microcystin-LR in CHO/OATP1B3 cells. (d) Cell viability based on microscopic observation after 24 h microcystin-LR exposure. Photographs were taken at a 10× magnification. 7806
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 20, 2009
There are currently a few assays available to test for microcystins, including screening assays such as the enzymelinked immunosorbent assay (ELISA), protein phosphatase inhibition assay (PPI), and mouse bioassay, which tests for the microcystin family as well as chromatography-based methods that can identify specific microsystin variants. However, there are limitations associated with these assays. Liquid chromatography coupled with mass spectrometry detection can identify and quantify specific microcystins at nanogram per liter concentrations depending on the limited commercial availability of reliable microcystin variant standards (17). However those methods are expensive, timeconsuming, and labor intensive as sophisticated equipment, experienced analysts, and sample extraction and clean up are required. Furthermore, the chromatography methods do not measure toxicity. Immunological procedures can reach a detection limit of 0.05 µg/L without sample concentration (4). However, ELISA results are not directly related to biological functions, and there is always the concern of cross reactivity and false positive results. The PPI assay is based on the fact that microcystins are potent inhibitors of protein phosphatases. The assay is sensitive to 0.05 µg/L without sample concentration (18). However, it is not specific for microcystins and it does not relate directly to in vitro toxicity. The mouse bioassay remains the standard method for assessing function and potency of cyanotoxins. It does not require sophisticated equipment and produces relatively quick results. However, it does not detect microcystins at low levels (especially in treated drinking water) with an assay sensitivity of about 3 µg/mL microcystins in water per mouse for lethality. Furthermore, it has severe limitations in routine laboratory application and environmental studies because most testing facilities do not have animal housing capabilities. To date, the best in vitro system available is based on the use of freshly isolated hepatocytes, which actively take up
FIGURE 5. Comparison of microcystin congeners cytoxicities. Curve 1 represents microcystin-LR dose-dependent cytotoxicity. IC50 ) 1.03 µg/L. Curve 2 represents microcystin-YR dose-dependent cytotoxicity. IC50 ) 2.24 µg/L. Curve 3 represents microcystin-RR dose-dependent cytotoxicity. IC50 ) 10.07 µg/L. Curve 4 represents microcystin-LF dose-dependent cytotoxicity. IC50 ) 0.14 µg/L. Curve 5 represents microcystin-LW dose-dependent cytotoxicity. IC50 ) 0.21 µg/L.
TABLE 1. Environmental Contaminants Tested Using a Real-Time Cytotoxicity Assay substance
experimental concentrations
2,4-D diclofop-methyl dinoseb diquat diuron ethalfluralin fenoxaprop-ethyl glyphosate dieldrin dimethoate endosulfan I endrin ethion lindane 4,6,8-trichlorophenol hexachlorobenzene
1-100 µg/L 1-100 µg/L 1-100 µg/L 1-100 µg/L 1-100 µg/L 1-100 µg/L 1-100 µg/L 1-100 µg/L 0.1-10 µg/L 10-100 µg/L 0.1-10 µg/L 10-100 µg/L 1-100 µg/L 0.1-10 µg/L 1-100 µg/L 0.1-1 µg/L
2,3,7,8-tetrachlorodi-benzo-p-dioxin PCB28 PCB138 pentachlorophenol bisphenol A carbofuran dioctyl phthalate 17β-estradiol
0.01-1 µg/L 100-10000 µg/L 100-10000 µg/L 10-100 µg/L 1000-10000 µg/L 100-10000 µg/L 100-10000 µg/L 10-7-10-13 M
microcystins. The LD50 for hepatocytes after 20 h incubation with microcystin-LR was reported to be 50 µg/L (19). The in vitro assay is relatively more sensitive than in vivo assays but still inadequate without sample concentration. It also requires live animals for tissue collection. Therefore, this assay needs improvement in sensitivity, stability, and consistency. To circumvent the existing problems and create an experimental model, we investigated immortalized cell lines as a potential platform to demonstrate the cytotoxic effect caused by microcystins. Because microcystins predominantly accumulate in the liver, we chose FL83B for the assay
type of substance
toxicity
herbicide herbicide herbicide herbicide herbicide herbicide herbicide herbicide insecticide insecticide insecticide insecticide insecticide insecticide industrial chemical metal refineries and agricultural chemical waste incineration waste chemicals waste chemicals wood preserver plasticizer soil fumigan fire retardant estrogen
kidney, liver, adrenal gland reproductive, carcinogenicity reproductive, carcinogenicity cataracts reproductive, carcinogenicity reproductive, carcinogenicity reproductive, carcinogenicity kidney, reproductive carcinogenicity reproductive, carcinogenicity reproductive, carcinogenicity liver reproductive, carcinogenicity liver, kidney, immunity liver liver, kidney, reproductive reproductive skin, thymus gland skin, thymus gland liver, kidney reproductive blood, neuro, reproductive reproductive, liver reproductive
development. We hypothesized that FL83B may mimic primary hepatocytes in response to microcystins. The results were not satisfactory because the assay sensitivity (100 µg/L) did not meet those achieved with freshly isolated hepatocytes (50 µg/L). Microcystins fail to travel across the cytoplasmic membrane of the immortalized hepatocytes (i.e., FL83B cells) and hepatocellular carcinoma cells (HepG2) as effectively as in freshly isolated hepatocytes. This observation indicates that hepatocytes likely change their gene expression pattern after recovery from liquid nitrogen preservation (20), which may result in the lack of transporter mechanism responsible VOL. 43, NO. 20, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7807
for microcystins getting into the cytoplasm. Therefore, modification to overexpress the related gene in cell lines was considered as a possible solution. OATP1B3 is found to be expressed exclusively in the liver. It plays important roles in various drug uptake mechanisms. Recently, it was demonstrated that OATP1B3 has the ability to transport microcystin in the Xenopus laevis oocyte expression system (12). Therefore, OATP1B3 was chosen as our target gene to establish an overexpression cell line for microcystin analysis. We indeed observed a sensitive, specific, and dose-dependent cytotoxic response in CHO/OATP1B3 cells but not in the CHO/WT control cells. This indicates that microcystin-LR can be transported into cells by OATP1B3, which results in acute cytotoxicity. Because OATP1B3 is exclusively expressed in the liver, our results may explain why microcystins accumulate predominantly in the liver and can cause severe organ-specific damage. Similar observations were made with four microcystin congeners, -YR, -RR, -LF, and -LW, but not the other environmental toxins tested, although some of them are known to be toxic to the liver as well. A possible explanation would be that those environmental toxins are taken up via a pathway other than OATP1B3 transporters. Further research is needed to prove this hypothesis. Our microcystin cytotoxicity assay sensitivity (
